The Advice Founders Ignore Until It’s Too Late

If you’ve spent time in NTE Pro (6,500+ startup ideas), you start to notice something quickly:

Ideas aren’t rare.
Effort isn’t rare.
Intelligence isn’t rare.

What’s rare is working on the right problem at the right time.

That’s why we’ve been coming back to Paul Graham’s essays. Not to quote them, but to use them. Today, let’s talk about another one that doesn’t get enough attention:

“What I’ve Learned from Users.”

It’s one of Paul’s most quietly devastating essays because it’s about what actually kills startups after the idea phase.

The Surprising Thing He Learned

After working with thousands of startups through Y Combinator, Paul realized something uncomfortable:

Most startups fail for the same reasons.
And founders almost always misunderstand which problems matter most.

Not because they’re careless.
Because they’re inside the maze.

When you’re building, everything feels urgent. Everything competes for attention. Everything sounds reasonable if you argue for it long enough.

That’s where things go wrong.

Founders Are Bad at Self-Diagnosis

Paul describes a pattern that shows up over and over:

Founders come in worried about fundraising.
The real issue is the product.

Founders come in worried about growth.
The real issue is that users don’t care enough yet.

Founders come in juggling five problems.
One of them will kill the company. Two don’t matter. One is annoying but survivable.

And they usually worry about the wrong one.

This isn’t a failure of intelligence.
It’s a lack of perspective.

Why Advice Sounds Wrong Until It’s Too Late

One of Paul’s sharpest insights in the essay is this:

Startup advice often sounds wrong when you need it most.

That’s not a bug, it’s the nature of startups.

They’re counterintuitive.
They don’t resemble school or jobs or anything most people have experienced before.
So when someone tells you something true but unfamiliar, it feels incorrect.

Founders don’t ignore advice because they’re stubborn.
They ignore it because they don’t believe it yet.

Experience is usually the only thing that convinces them.

Patterns Matter More Than Genius

This is where things get interesting.

Paul points out that once you’ve seen enough startups, the problems stop being mysterious. They repeat. The details change, but the structure doesn’t.

That’s a huge advantage if you can access it.

This is why studying outcomes beats debating opinions.

It’s also why we believe tools like WhoFiled should exist.

Instead of asking, “What should I do?”
You can ask, “What actually worked for companies in situations like this?”

When you see thousands of real examples — what got funded, how it was framed, who backed it, and when, you start to diagnose faster and more accurately.

Not because you’re smarter.
Because you’re less blind.

The Problem Behind the Problem

Paul describes another failure mode that hits founders hard:

Even when founders know their problems, they misjudge their relative importance.

They’ll spend weeks polishing something that doesn’t matter while ignoring the one issue that determines survival.

This is where early-stage focus breaks down.

And it’s why in NTE Zero to One, the biggest unlock often isn’t a new idea, it’s clarity about what actually matters first.

Why Users and Investors Are Better Mirrors Than Intuition

Paul’s essay is called What I’ve Learned from Users for a reason.

Users don’t care about your story.
Investors don’t care about your intentions.
They respond to value.

When founders struggle, it’s often because they’re listening to themselves instead of listening to reality.

Capital flow is one form of that reality.
User behavior is another.

WhoFiled makes capital legible.
NTE Pro makes idea patterns visible.

Together, they shorten the feedback loop founders usually have to learn the hard way.

The Uncomfortable Takeaway

Paul doesn’t say this outright, but it’s implied everywhere:

Most founders don’t fail because they didn’t try hard enough.
They fail because they worked very hard on the wrong thing for too long.

The earlier you can diagnose correctly, the faster everything else improves:
Ideas sharpen.
Focus improves.
Execution speeds up.

Why We Keep Coming Back to Paul Graham

We’re sticking with Paul Graham’s essays because they don’t motivate you.

They recalibrate you.

They don’t tell you what to build.
They teach you how to notice when you’re lying to yourself.

And if you’re honest about that, everything else gets easier.

“What I’ve Learned from Users” is worth reading slowly.

If you want to see those lessons play out in real time, WhoFiled gives you visibility into how companies are actually evaluated, funded, and framed and NTE Pro turns those patterns into ideas you can act on.